Development of optimization-based simulation tool for trajectory planning in planar arm reaching after stroke

Rehabilitation of people with upper extremity motor deficits is typically focused on relearning of motor abilities and functionalities requiring interaction with physiotherapists and/or rehabilitation robots. In a point-to-point movement training, the trajectories are usually arbitrarily determined without considering the motor impairment of the individual. In this paper, we report on development of an optimal control model based on arm dynamics enabling also incorporation of muscle functioning constraints (i.e. simulation of muscle tightness) to study optimal trajectories for planar arm reaching movements. In preliminary tests of the developed model we first tested the ability of the minimum joint torque cost function to replicate trajectories obtained in previously published experimental trials done by neurologically intact subjects, and second, we explored optimal trajectories when muscle constraints were modeled. The trajectories generated by the model without implemented muscle constraints show considerable similarity as compared to the experimental data, while muscle constraints considerably changed optimal trajectories.